关键词: 云计算, 服务级别协议（SLA）, 资源分配, 基于模拟退火的粒子群优化算法（SA-PSO）

Abstract: For the problem, how to efficiently allocate the resource to ensure the service provider to make the maximize profit in cloud computing, this paper proposes a dynamic cloud resource allocation policy based on Service Level Agreement （SLA）, by considering the SLA properties, containing computing capacity, network bandwidth and data storage as optimization parameters. A kind of mapping model between service request and resource allocation is given, at the same time, the corresponding utility function is designed. An improved hybrid Particle Swarm Optimization algorithm based on Simulated Annealing （SA-PSO） is analyzed. Experimental results show that the SLA-based parameters SA-PSO has better global optimality. It can obtain higher benefits than other algorithm under the same condition.

Key words: cloud computing, Service Level Agreement（SLA）, resource allocation, Particle Swarm Optimization algorithm based on Simulated Annealing（SA-PSO）